Comments for phycoweb blog

Comment on Visualizing niche conservatism by heroen

heroen — Mon, 16 Jul 2012 12:11:54 +0000

Hi Oli, that’s actually what we did here. The order of the species in the rows/columns corresponds to their order of appearance in a ladderized rooted phylogram. So closely related species are, on average, closer to each other in the rows/columns, the inevitable exception being when adjoining rows/columns represent species from different deeply branching clades. It would have been informative to add the phylogeny on the plot to show the species’ relatedness.

Comment on Visualizing niche conservatism by Olivier

Olivier — Mon, 16 Jul 2012 11:29:53 +0000

How about presenting the niche similarities by incorporating the phylogenetic structure of the species? You could order the species according to their phylogenetic relatedness (~ somewhat similar to heatmaps of gene families to visualise expression data).

Comment on No name, new game by Roderic Page (@rdmpage)

Roderic Page (@rdmpage) — Wed, 16 May 2012 07:04:32 +0000

Several things seem to be coming out of this discussion. The first is that sequence data make taxonomy computable. We can assign specimens to taxa with an associated measure of confidence, and we can do this on a massive scale (e.g., DNA barcoding). A challenge is connecting sequence data with existing taxonomy (assuming that we want to do that, perhaps in some groups we might argue it would be better to start afresh). Given the amount of “dark taxa” we are going to struggle to do this using names alone. Specimens are another way to do this, but this requires that we have vouchers, and ideally vouchers that have globally unique identifiers so that we can link sequences, specimens, and data on those specimens (e.g., papers that cite those specimens as part of a taxonomic study). We are some way from being able to do this, but perhaps the rise of dark taxa will focus people’s minds on this task.

Comment on No name, new game by Gary Saunders

Gary Saunders — Fri, 04 May 2012 12:43:53 +0000

Interesting and controversial topic – here are my own dogmatic opinions. Does a species need a name (?) – absolutely! Must it be an ICBN-compatible binomial (?) – absolutely not! A species identifier could be based on any form of alpha and/or numeric label so long as there is global adoption of that system. It could even be a DNA sequence from an agreed upon marker. My preference would be that every species receive a unique number from its initial entry into a database. In the ideal world that database would be an internationally accredited/ recognized genetic database, with the sequence data diagnostic of that species rather than used as the unique identifier itself. BOLD loosely does this by assigning collections to genetic bins, each with a unique identifier (Cluster number). Taxonomists could then go about their business of assigning names, correcting names, transferring species, etc., under the Linnaean binomial system (let’s face it, Claudea elegans is far more eloquent than, for example, “TH45238ug437”). Regardless to the nomenclatural and taxonomic chaos that may ensue, that unique number would remain attached to that ‘species’ thus making it very easy for non-taxonomists to compare data between studies, databases, etc., without having to be fully conversant with the taxonomic literature. This is a pragmatic approach that bridges the romance of the pre-molecular age of ‘Linnaean’ taxonomic discovery with the contemporary DNA-driven age of discovery.

Comment on No name, new game by Marie Pazoutova

Marie Pazoutova — Sun, 29 Apr 2012 22:32:57 +0000

Yes, indeed! The mess in Genbank is like the rise of “wild capitalism” in Mid/Eastern Europe right after the fall of communist regimes: the process bursted spontaneously and was quicker than our ability to recognize we need some rules and regulations…

I think we do need the binomials. More than 200 years of science was done with them. I guess that leaving the names would cause more problems than keeping them.
The problem is not whether to leave them or not, but how to reconcile the traditional system with the modern (that will be out-of-date again in few decades!). And, subsequently, how to set the ultimate (ultimate? for how long?) species concept and decide at which level we allow the splitting of cryptic species or – the other way round – unifying and merging. And also: are we able to set thes rules to be applied universaly accros the whole eukaryotic (at least) web of life (leaving the bacteriologists aside)?

Comment on No name, new game by Paul Gabrielson

Paul Gabrielson — Thu, 19 Apr 2012 23:32:55 +0000

To name or not to name, that is the question. As an old school morpho-taxonomist, that part of me says, “of course we have to use validly published names”; as a practicing molecular systematist who currently is surprised and inundated by the number of undescribed macro-marine algal taxa in the well-worked NE Pacific flora, I am more sympathetic to alternatives to the binomial nomenclature I still adhere to. But I have become increasingly frustrated with slow pace of that work, as I am overwhelmed by how much we don’t know. At present, I am focused on coralline algae, a group in the NE Pacific I thought would benefit from the fresh approach of DNA sequencing, and that has proved to be true. But I now find myself hoping that a specimen I have sequenced will match a sequence already in my database and not be new. There are too many undescribed species to be characterized, and we are just scratching the surface.

As Samyn & De Clerck (2012) have pointed out and reinforced by Heroen’s red algal graph, the number of “dark taxa” is increasing steadily, and, I believe, will increase at a greater rate in the next 5-10 years with intensive collecting coupled with bar-coding. But these so-called “dark taxa” at least have an identifier, a DNA sequence. For these dark taxa not having a generic name and specific epithet is a sin of omission that is easily rectified over time by providing a name. What is more troubling, to this traditional taxonomist, is the sin of commission, the assigning of a validly published name to a sequence when there is no clear link to the type specimen on which the name is based. There are many of these in GenBank, and they are far more misleading than a generic name followed by sp. Equally troubling is that many of these named sequences lack a voucher specimen, which means their identity cannot be verified. It is better to not name a sequence in a public database than to provide a name about which one is not certain. Correcting these is more difficult and time consuming than assigning unknown sequences a name, once the identity of the sequence is known. Let’s err in naming with sins of omission rather than those of commission.

Paul Gabrielson
Herbarium Research Associate
University of North Carolina, Chapel Hill
drseaweed@hotmail.com